Back to skill

Security audit

WalletPrint

Security checks for vulnerabilities and agentic risk

Overview

WalletPrint is a coherent transaction-risk scoring skill that uses an external hosted API as advertised, but users should understand that wallet and transaction metadata may be sent and stored.

Install only if you are comfortable sending proposed transaction metadata to WalletPrint's hosted service. Use scoped per-agent keys instead of master keys in agents, keep WALLETPRINT_BASE_URL pinned to the official HTTPS API unless intentionally testing, avoid putting sensitive details in feedback notes, and secure any webhook receiver before using it for approval decisions.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (2)

T09 · Insecure Skill Coding Practices

Warning
Location
src/client.ts:15
Finding

Unrestricted API Base URL Can Expose API Credentials and Transaction Metadata

Content
View full analysis
{ return this.request("POST", "/v1/score", request); } async submitFeedback(request: FeedbackRequest): Promise { return this.request("POST", "/v1/feedback", request); } async health(): Promise<{ status: string }> { return this.request<{ status: string }>("GET", "/health"); } private async request( method: "GET" | "POST", path: string, body?: unknown, ): Promise { const response = await this.fetchImpl(`${this.baseUrl}${path}`, { method, headers: { "content-type": "application/json", "x-api-key": this.apiKey, }, body: body ? JSON.stringify(body) : undefined, }); ``` The associated normalization function performs no security validation: ```typescript export function normalizeBaseUrl(baseUrl: string): string { return baseUrl.replace(/\/+$/, ""); } ``` ### Technical Analysis The client treats `baseUrl` as trusted after merely removing trailing slashes. It does not parse the value, enforce HTTPS, reject embedded credentials, or restrict the destination to an expected API origin. Every request sends the configured API key in the `x-api-key` header. Scoring requests also contain financial and behavioral metadata, including: - Sender wallet address - Recipient wallet address - Blockchain network - USD-normalized transaction value - Asset and contract metadata - Optional platform, environment, and agent identifiers Although sending this information to a hosted scoring service is necessary for the declared function ...[truncated 1393 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
docs/approval-flow.md:88
Finding

Reference Approval Webhook Handler Accepts Unauthenticated Alert Payloads

Content
View full analysis
{ const payload = req.body; const reasons = payload.reason_codes .map((r: { label: string; detail: string }) => `• ${r.label}: ${r.detail}`) .join("\n"); await fetch(process.env.SLACK_WEBHOOK_URL!, { method: "POST", headers: { "content-type": "application/json" }, body: JSON.stringify({ text: `WalletPrint flagged a ${payload.band} transaction (score ${payload.score})`, blocks: [ { type: "section", text: { type: "mrkdwn", text: `*${payload.band.toUpperCase()}* score ${payload.score}\n${reasons}`, }, }, { type: "section", fields: [ { type: "mrkdwn", text: `*Wallet*\n\`${payload.wallet_address}\`` }, { type: "mrkdwn", text: `*To*\n\`${payload.to_address}\`` }, { type: "mrkdwn", text: `*Amount*\n$${payload.value_usd} ${payload.asset}` }, { type: "mrkdwn", text: `*Chain*\n${payload.chain}` }, ], }, ], }), }); res.sendStatus(200); }); ``` ### Technical Analysis The documented reference handler accepts any request reaching `/walletprint/webhook` and trusts `req.body` without: - Verifying a cryptographic webhook signature - Authenticating the sender - Checking a timestamp or replay identifier - Validating the payload against a strict schema - Applying request-size or rate restrictions It then forwards attacker-controlled fields to a Slack operational channel. Because this example is presented as part of a human approval workflow, developers may copy it into production and unintentionally expose an unauthenticated alert-inject ...[truncated 1420 chars]
Remediation
View remediation
{ const signature = req.header("walletprint-signature"); const timestamp = req.header("walletprint-timestamp"); if (!verifyWalletPrintSignature(req.body, signature, timestamp)) { return res.sendStatus(401); } const payload = webhookSchema.parse(JSON.parse(req.body.toString("utf8"))); if (await wasEventPreviouslyProcessed(payload.screened_transaction_id)) { return res.sendStatus(409); } // Forward the validated event to the approval channel. return res.sendStatus(200); }, ); ``` ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
Findings (43)

YARA rule 'agent_skill_credential_exfiltration_webhook': AI agent skill credential harvesting followed by webhook or external exfiltration [agent_skills]

Critical
Category
YARA Match
Confidence
85% confidence
Finding

YARA rule matched a known malware signature (reverse shell, backdoor, ransomware, C2 framework, or info stealer).

Content

Scanner excerpt · docs/approval-flow.md (reported line 25)May include surrounding context.

ever executes approve/deny logic. You own the approval UX and the final signing decision.

Configure webhooks

Production integrators can register a webhook URL and choose which bands trigger delivery:

bash
curl https://walletprint.up.railway.app/v1/webhook \
  -X PATCH \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_PRODUCTION_API_KEY" \
  -d '{
    "webhook_url": "https://hooks.slack.com/services/T000/B000/XXXX",
    "webhook_bands": ["medium", "high"]
  }'

Set "webhook_url": null to disable webhooks.

Default bands: medium and high. Low-band scores are not delivered unless you include "low" in webhook_bands.

Webhook payload

When a scored transaction matches your configured bands, WalletPrint POSTs JSON to your webhook URL:

json
{
  "event": "transaction.flagged",
  "timestamp": "2025-06-20T12:00:00.000Z",
  "screened_transaction_id": "9ffc282b-ac2b-4249-999a-1b68c8a91756",
  "score": 65,
  "band": "medium",
  "reason_codes": [
    {

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · .gitignore (reported line 3)May include surrounding context.

text
node_modules/
dist/
.env
.env.*
!.env.example
*.log

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
80% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · docs/api.md (reported line 226)May include surrounding context.

}

text

## `DELETE /v1/agent-keys/:agent_key_id`

Revoke an agent key. **Master key required.**

Known Vulnerable Dependency: nanoid==3.3.12 — 2 advisory(ies): CVE-2026-67214 (nanoid: non-secure generators can loop indefinitely with negative size); CVE-2026-67213 (nanoid: custom generators can loop indefinitely when size is zero)

High
Category
Supply Chain
Confidence
93% confidence
Finding

nanoid 3.3.12 is reported as vulnerable to infinite-loop denial of service conditions in non-secure or custom generator scenarios when given invalid sizes. As a transitive dev dependency this does not imply direct exploitability, but if build or tooling code accepts attacker-controlled size parameters it could hang local tooling or CI jobs.

Content

No source excerpt is available for this finding.

Known Vulnerable Dependency: postcss==8.5.15 — 2 advisory(ies): CVE-2026-69153 (PostCSS: incomplete fix of GHSA-6g55-p6wh-862q — attacker-controlled sourceMappi); CVE-2026-73646 (PostCSS: Path Traversal in Previous Source Map Auto-Loading (sourceMappingURL) l)

High
Category
Supply Chain
Confidence
95% confidence
Finding

postcss 8.5.15 is flagged for source map handling issues including path traversal / unintended file access when processing attacker-controlled CSS or sourceMappingURL data. Because PostCSS is commonly exercised in build pipelines, this could expose sensitive files or cause unsafe filesystem reads in developer or CI environments if untrusted content is processed.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · README.md (reported line 49)May include surrounding context.

Create scoped keys for individual agent wallets (server-side, using your master key):

typescript
const response = await fetch("https://walletprint.up.railway.app/v1/agent-keys", {
  method: "POST",
  headers: {
    "content-type": "application/json",

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · docs/getting-started.md (reported line 91)May include surrounding context.

Create scoped keys for individual agent wallets (server-side, using your master key):

typescript
const response = await fetch("https://walletprint.up.railway.app/v1/agent-keys", {
  method: "POST",
  headers: {
    "content-type": "application/json",

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · README.md (reported line 251)May include surrounding context.

WalletPrint is a signal layer — not an approval UI. Register a webhook to receive medium/high band alerts and wire them into Slack, email, or your own review flow:

bash
curl https://walletprint.up.railway.app/v1/webhook \
  -X PATCH \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_PRODUCTION_API_KEY" \

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill instructs agents to send wallet addresses, recipient addresses, chain, asset, and transaction amount details to a third-party WalletPrint API for risk scoring, but it does not clearly disclose this data transfer or its privacy implications. In an agent context, this can cause sensitive financial metadata to be exfiltrated to an external service without informed user consent, which is especially risky when used automatically before signing transactions.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · docs/api.md (reported line 26)May include surrounding context.

Returns service health.

bash
curl https://walletprint.up.railway.app/health

Response:

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The documentation explicitly states that screened transactions, wallet history, cluster signals, and human feedback are persisted and exportable, but it does not pair that with clear privacy, retention, or data-governance guidance. In an agent integration context, this can cause operators to send sensitive financial activity and reviewer annotations to a third-party service without understanding storage duration, downstream use, or compliance implications.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
88% confidence
Finding

This scoring endpoint sends wallet addresses, recipient addresses, transaction value, asset details, and optional contextual metadata to an external hosted service. In this specific skill, the danger is elevated by nearby documentation stating that production submissions are persisted and used to build baselines, meaning integrators may disclose sensitive transaction intelligence without prominent consent or minimization guidance.

Content

Scanner excerpt · docs/api.md (reported line 86)May include surrounding context.

Example with marketplace metadata:

bash
curl https://walletprint.up.railway.app/v1/score \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · docs/api.md (reported line 163)May include surrounding context.

Create a per-agent API key. Master integrator key required. Production keys only.

bash
curl https://walletprint.up.railway.app/v1/agent-keys \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_MASTER_KEY" \
  -d '{

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · docs/api.md (reported line 203)May include surrounding context.

List agent keys for your integrator. Master key required.

bash
curl https://walletprint.up.railway.app/v1/agent-keys \
  -H "x-api-key: YOUR_MASTER_KEY"

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The webhook section explains how to forward flagged transaction events to an arbitrary URL, but it does not clearly warn that transaction-related data will leave the platform and be delivered to a third party controlled by the integrator. This omission increases the risk of accidental disclosure to insecure, misconfigured, or unexpected endpoints, especially in automated agent workflows.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
84% confidence
Finding

Triggering a webhook test causes the platform to send a payload to a configured external URL, which is an outbound transmission of transaction-like event data to third-party infrastructure. Because the docs do not foreground the privacy and trust implications of that delivery, operators may test against insecure or unintended endpoints and leak operational metadata.

Content

Scanner excerpt · docs/api.md (reported line 320)May include surrounding context.

Send a test webhook payload to your configured URL. Master integrator key required. Production keys only.

bash
curl https://walletprint.up.railway.app/v1/webhook/test \
  -X POST \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_MASTER_KEY" \

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · docs/approval-flow.md (reported line 20)May include surrounding context.

Production integrators can register a webhook URL and choose which bands trigger delivery:

bash
curl https://walletprint.up.railway.app/v1/webhook \
  -X PATCH \
  -H "content-type: application/json" \
  -H "x-api-key: YOUR_PRODUCTION_API_KEY" \

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The documentation explicitly encourages forwarding wallet addresses, recipient addresses, transaction amounts, and related risk metadata to third-party services such as Slack and email providers. While this appears intended for operational alerting, it creates a real privacy and data-sharing risk because sensitive transaction context is transmitted outside the primary system without any warning, minimization guidance, or verification controls.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The README explicitly says the example 'immediately submits a confirmed_benign label' but does not warn that this sends externally meaningful feedback data about a transaction. Silent or poorly disclosed transmission of labels can leak sensitive operational judgments, create privacy/compliance issues, and cause incorrect irreversible feedback if triggered automatically.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

This code sends transaction-derived metadata, including wallet address, recipient address, asset, contract address, value, and optional context, to an external scoring client before submission. Even if intended for fraud/risk screening, the file provides no consent, minimization, or visibility guarantees, so sensitive financial activity may be disclosed to a third party unexpectedly.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The wrapper performs a risk-scoring call before signing, but it never enforces the result and always proceeds to sendTransaction(transaction). This creates a security-gap vulnerability because integrators may reasonably rely on the wrapper for protection based on the API/docs, while malicious or high-risk transactions are still executed unchecked.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

This code constructs a score request containing walletAddress, chain, transaction details, asset, transactionType, and context, then sends it via client.score(). There is no confirmation prompt, logging, or explicit user-facing disclosure in this file that potentially sensitive wallet and transaction metadata will be transmitted to an external scoring service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · test/sdk.test.ts (reported line 47)May include surrounding context.

ts
expect(result.score).toBe(27);
    expect(fetchImpl).toHaveBeenCalledWith(
      "https://api.example.com/v1/score",
      expect.objectContaining({
        method: "POST",
        headers: expect.objectContaining({

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

The quick-start example shows client.score() sending wallet address, recipient, asset, and transaction amount to a remote service, but the immediate section does not warn users that this data is transmitted off-system. Although later sections mention hosted API calls and some storage behavior, the primary usage path lacks a clear user-facing warning near the data-sharing action.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The feedback flow sends screened transaction IDs, outcome labels, and arbitrary free-form notes to the external service, but the skill does not warn users that these post-transaction details leave the local environment. Free-form notes are particularly risky because operators may include sensitive operational, financial, or identifying information that is unnecessarily disclosed to a third party.

Content

No source excerpt is available for this finding.